Analysis of Chemical Characteristics of Fluoride-Rich Waters in Deeper Coastal Aquifers of Alappuzha District, Kerala State, South India for Domestic and Irrigation Purposes

Abstract

The population of the coastal district of Alappuzha depends mainly on groundwater for domestic needs. The poor quality of surface and shallow groundwater conditions, deeper aquifers are heavily exploited by the population. Analysis of chemical characteristics of Fluoride-rich waters in deeper coastal aquifers of the area is very much required for testing its uses for various needs. The concentration of fluoride in groundwater above the permissible limit WHO (World Health Organization) has been reported in many Tehsils of the district. The presence of amphiboles (Hornblende), fluorapatite, fluorite, and micas may be responsible for the F⁻ in groundwater (Todd, Groundwater Hydrology, John Wiley, New York, 2005). Fluoride contamination above the maximum permissible limit may cause mottling of teeth, dental caries, dental fluorosis, and skeletal fluorosis (Karanth, Ground water assessment, development and management. Tata McGraw-Hill Ltd, New Delhi, 1987). The water quality studies were done by Sen (J Am Stat Assoc 63:1379–1389, 1968), Handa (Groundwater 13:275–281, 1975), Seyhan et al. (Water Resour Res 21:1010–1024, 1985), Razack and Dazy (J Hydrol 114:371–393, 1990), Malini et al. (Int J Environ Stud 60:87–98, 2003), Jeevanandam et al. (Environ Monit Assess 132:263–274, 2006), Olobaniyi and Owoyemi (Afr J Sci Technol Sci Eng Ser 7:73–81, 2006), and many others. The present work on the analysis of chemical characteristics of Fluoride-rich waters in deeper coastal aquifers of Alappuzha District, Kerala State, South India for domestic and irrigation purposes is an attempt to know the variables controlling the water quality of deeper groundwater reservoirs using statistical tools.

1 Introduction

The population of the coastal district of Alappuzha depends mainly on groundwater for domestic needs. The poor quality of surface and shallow groundwater conditions, deeper aquifers are heavily exploited by the population. Analysis of chemical characteristics of Fluoride-rich waters in deeper coastal aquifers of the area is very much required for testing its uses for various needs. The concentration of fluoride in groundwater above the permissible limit WHO (World Health Organization) has been reported in many Tehsils of the district. The presence of amphiboles (Hornblende), fluorapatite, fluorite, and micas may be responsible for the F⁻ in groundwater (Todd 2005). Fluoride contamination above the maximum permissible limit may cause mottling of teeth, dental caries, dental fluorosis, and skeletal fluorosis (Karanth, 1987). The water quality studies were done by Sen (1968), Handa (1975), Seyhan et al. (1985), Razack and Dazy (1990), Malini et al. (2003), Jeevanandam et al. (2006), Olobaniyi and Owoyemi (2006), and many others. The present work on the analysis of chemical characteristics of Fluoride-rich waters in deeper coastal aquifers of Alappuzha District, Kerala State, South India for domestic and irrigation purposes is an attempt to know the variables controlling the water quality of deeper groundwater reservoirs using statistical tools.

2 Study Area

Alappuzha is the smallest district in Kerala having an area extent of 1414 km² and it accounts for 3.64% of Kerala. The study area lies between N lat. 09°05′ and 09°54′ and E long. 76°16′ and 76°46′ falling in the Survey of India degree sheet No. 58 C (Fig. 1). The Lakshadweep Sea borders the area on the western side. According to the 2011 census population of the area is 2121943 with a density of 1501 persons/km².

Fig. 1

Location map of Alappuzha District, Southern Kerala, India

The soil types reported in the area are alluvium (coastal & riverine), brown hypidimorphic soil, and lateritic. Paddy (Oryza sativa) is the principal crop cultivated in an area (344.74 km²) followed by coconut (Cocos nucifera) and banana (611.0 km²). The important plantation crops are rubber, cashew, pepper (130 km²), and tapioca (03.0 km²). The discharges of the four major river systems emptying into the Vembanad Lake pass through Thanneermukkom where a barrage was constructed to prevent saltwater intrusion. The barrier, to a large extent, has effectively prevented the intrusion of saline waters into the Kuttanad region. The SW monsoon (during June–September) and NE monsoon (from October to December) contribute 60.3 and 20.9% of the annual rainfall respectively and 18.8% of the rainfall received from January to May. A major part of the area is classified as a coastal plain represented by various depositional features formed under marine, fluvio-marine, and fluvial environments. The Kuttanad area is lying below msl (1–2 m below msl) and the remaining area has an altitude below 6 m. Alappuzha district is mainly drained by the Pamba River and its two tributaries—Achenkovil and Manimala. A major part of the district is covered by sedimentary formations ranging in age from Eocene to Recent overlying the Precambrian crystallines. The stratigraphic sequence of the area is compiled (Table 1).

Table 1 Stratigraphic sequence of the area

3 Methodology

There were a total of 75 groundwater samples collected from deeper aquifers (tube wells) from Cherthala, Kuttanad, and Ambalapuzha tehsils of Alappuzha district. The pre-monsoon (PRM) samples (First week of April) were collected in polyethylene bottles in 2011, 2012, and 2013 and analysed for major chemical quality parameters were determined as per the standard guidelines (APHA 2017). The EC and pH metres were used for the spot measurements EC and pH. The hydrochemical parameters were tested for Electro-Neutrality (E.N, %) (Huh et al., 1998) and were within ± 5%. The geochemical characterization and evolution groundwater of deep aquifers in the water samples for the pre-monsoon samples of 2013 have been deeply discussed. Various statistical treatments especially Factor analysis carried out by using SPSS 16.0 for Windows.

4 Results and Discussion

The fluoride contamination, variation of fluoride trend over the years, processes involved in the concentration of fluoride and the remedial measures for mitigating the adverse effects of higher fluoride in the deeper aquifers of the area are discussed. A high concentration of fluoride has been observed in many villages of the Alappuzha district. The present study includes water sampling in 25 tube wells of different villages from those areas with high levels of fluoride during the 2011, 2012, and 2013 PRM (First week of April) periods. The geological cross-section along the coastal tract of Alappuzha/geological profile has been prepared (Fig. 2) to know the aquifer geometry of the terrain. The hydrogeological details (piezometric head, depth, yield, zone taped, etc.) of the inventoried key wells are compiled (Table 2) and the geological Profile along the coastal stretch of the Alappuzha district is given in Fig. 2. The aquifers from which water is being extracted are sandstones with pebbles and gravel beds belonging to Vaikom beds and are confined/pressure aquifers. The water chemistry of the observation wells is compiled (Table 3).

Fig. 2

Geological profile along Coastal Alappuzha and nearby areas (Anon 2016)

Table 2 Hydrogeological details of observation wells under study

Table 3 Results of chemical analysis of selected ground water samples in Alapuzha district

The fluoride concentration during the study period varied between 0.44 mg/L (Arookutty) and 2.48 mg/L (Cherthala Town). The lowest and the highest fluoride concentrations were reported at Arookutty and Cherthala towns respectively during different periods of sampling. The correlation of Ca²⁺, Mg²⁺, total hardness_, K⁺, Na⁺, pH and EC with fluoride has been studied for the water samples of 2011, 2012, and 2013. The study revealed that Ca²⁺, Mg²⁺, and H_T have a negative correlation with fluoride and is owing to a little solubility of the fluorides of these ions (Hem 1991; Hounslow 1995). The alkali metal ions viz. K⁺ and Na⁺, pH and EC showed a positive correlation with fluoride content. Table 4 shows the summarized results of water classification.

Table 4 Summarised results of water classification

The chemical analysis data for the different periods have been used for graphical representation in the Piper diagram and the plots are compiled (Fig. 3). The analysis of the diagrams revealed that there is a strong correlation among Na⁺ and Cl⁻, which in turn is indicative of the marine signatures in the area and the hydrochemical facies of the samples are compiled (Table 4).

Fig. 3

Piper diagrams of water samples for different periods

The EC values of the samples varied from 334 to 4021 µs/cm at 25 °C and pH 6.92 to 8.70. The average values for pH varied from 7.25 to 7.81, indicating that the waters were slightly alkaline during the seasons in three years. In 3 out of 25 locations, the pH values are more than the desirable limit (8.5) of drinking water guidelines of WHO (1996) standard. The TDS ranged from 187 to 2233 mg/L and average values of 672.2–811.48 mg/L. In 62.7% of the samples, TDS exceeded the desirable limit (500 mg/L) as per WHO guidelines. As per Table 3, it is observed that the order of abundance is Na⁺ ≥ Ca⁺⁺ ≥ Mg⁺⁺ ≥ K⁺ (cations) and in 97.5% of samples Na⁺ exceeded (75 mg/L). But Ca⁺⁺, Mg⁺⁺, and K⁺ are much lesser than the maximum level. The concentration of Ca⁺⁺ ranged from 16 to 73 mg/L, Mg⁺⁺ from 5.0 to 43 mg/l and K⁺ ranged from 2.1 to 11.8 mg/L. The abundance of anions is Cl⁻ ≥ HCO₃ ≥ SO₄⁻ and the Cl⁻ varied between 45 and 1092 mg/l (average from 381 to 405.24 mg/L) and in 80% of samples the Cl⁻ exceeded the desirable limit (200 mg/l), but the sulphate varied between 4 and 27 mg/L which are less than the prescribed one by WHO (1996) (400 mg/L). The concentration of HCO₃⁻ varied between 67 and 489 mg/L (with average values from 139.96 to 233.76 mg/L). The nitrate varied from 1.3 to 23.6 mg/L (mean values from 9.38 to 10.74 mg/L). In two tube wells, the nitrate value exceeded the desirable limit (20 mg/L). In the case of Fe⁺⁺, the Fe⁺⁺ varied between 0.1 and 0.86 mg/L and the mean values were from 0.28 to 0.31 mg/L. In 42.7% of the samples, the Fe⁺⁺ exceeded the desirable limit (0.3 mg/L). The H_T varied between 88 and 498 mg/l with mean values from 18.76 to 231.36 mg/L and in 52.3% of cases the total hardness exceeded the permitted limit of 300 mg/L. The fluoride varied between 0.44 and 2.52 mg/L with mean values from 1.64 to 1.79 mg/L. Also, it is to be noted that in 14 out of 25 wells, fluoride exceeded 1.5 mg/L and in 93.3% of cases it exceeded the desirable limit of 1.0 mg/L. The parameters TDS, HCO₃, T_H, SO₄⁻, Ca⁺⁺, Mg⁺⁺, NO₃⁻, K⁺, and F⁻ were shown an increasing trend over the three years, but pH and Fe⁺⁺ showed a decreasing trend. The hydrochemical parameters pH, TDS, T_H, Cl⁻, Fe⁺⁺, Na⁺, NO₃⁻, and fluoride exceeded the desirable limit of drinking water and the statistical summary is compiled (Table 5).

Table 5 Statistics of hydrochemistry of tube well water samples

The correlation between hydrochemical parameters is compiled (Table 6). The pH is significantly negatively correlated with sulphate and nitrate. The TDS is significantly positively correlated with EC, SO₄, Cl⁻, Na⁺, K, and fluoride; significantly negatively correlated with T_H and Mg⁺⁺. The parameter increased together with T_H are Fe⁺⁺, Ca⁺⁺, Mg⁺⁺, and NO₃⁻. The parameters which are significantly negatively correlated with T_H are EC, TDS, SO₄⁻, Cl⁻, Ca⁺⁺, Na⁺, K⁺, and F⁻. The parameters which are significantly correlated with Cl⁻ are EC, TDS, TH, SO₄⁻, Mg⁺⁺, Fe⁺⁺, and Na⁺ of which T_H and Mg⁺⁺ are negatively correlated. The parameters which show significant correlation with Fe⁺⁺ are T_H, Cl⁻, Ca⁺⁺, NO₃⁻, K⁺, and F⁻ of which K⁺ and F⁻ are negatively correlated. The parameter fluoride is significantly positively correlated with TDS, HCO₃⁻, and K⁺; significantly negatively correlated with T_H, Fe⁺⁺, Ca⁺⁺, Mg⁺⁺, and NO₃⁻.

Table 6 Correlation matrix of hydro chemical parameters Karl Pearson’s Correlation coefficient (p-value)

The water quality parameters were grouped using factor analysis. In factor analysis, the correlation matrix of variables was generated and factors were extracted using principal component analysis, rotated by varimax with Kaiser Normalization. Factor loadings, communalities for each variable, percentage of the variance of each factor, and cumulative percentage of the variance of the three-factor scores are compiled (Table 7).

Table 7 Factor loadings and commonality of the variables after varimax rotation

In 2011, commonalities of all the ions except Fe²⁺ are more than 0.7 and in 2012 except HCO₃⁻ the commonalities of Cl⁻, Fe²⁺, Mg²⁺, NO₃⁻, and K⁺ are more than 0.7 and in 2012, except Fe⁺⁺ the commonalities of SO₄⁻, Mg⁺⁺, NO₃⁻, Na⁺, and K⁺ are more than 0.7. The higher Eigenvalues for the first three factors ranged from 66.32% to 69.79% of the variance. Factor 1 which is associated with the variables EC, TDS, Cl⁻, and Na⁺ explains 27.2–37.1% of the variances. This factor reflects the signature of natural water recharge, and rock–water relations and salinity. Factor 2 accounts for 22.4–25% of the variance and Factor 2 has high loadings with T_H, Fe⁺⁺, Ca⁺⁺, Mg⁺⁺, HCO₃⁻, K⁺, and F⁻. This factor reflects the signature of cations/weathering. Factor 3 accounts for 10.2% to 13.2% of the variance, which is connected by pH, SO₄⁻, and NO₃⁻. This factor may be due to the effect of lithological sources and agricultural fertilizers.

To know the association between fluoride and other hydrochemical parameters, stepwise regression analysis has been utilized. In stepwise regression analysis, fluoride is taken as the dependent and other parameters as independent variables. The model identified the significant parameters and eliminated other parameters from the model. The model identified that Ca²⁺, NO₃, TDS, and Fe²⁺ were significant. The R square measures the proportion of variance of fluoride explained by the independent variables together, which is obtained as 0.696. That is, 69.6% of the variation in fluoride is explained by the variables Ca⁺⁺, NO₃⁻, TDS, and Fe⁺⁺. The ANOVA table (Table 8) shows the model is significant at a 1% level of significance (F-value = 40.144 and sig. = 0.000). It can be seen from the table of coefficients (Table 9) that Ca²⁺, NO₃⁻, and Fe²⁺ are significant and negative but the TDS is positive. The collinearity statistics checked the uncorrelation condition of independent variables, which are under the permissible level. The fluoride contents beyond the guideline values of WHO have been found in 84% of the samples of deeper aquifers.

Table 8 ANOVA

Table 9 Coefficients

The evolution of the groundwater samples of the area under study (PRM 2013) has been attempted. As per Soltan (1998) groundwater may be Normal Chloride type (Cl⁻ < 15 meq/L), Normal Sulphate type (SO₄²⁻ < 6 meq/L), and Normal Bicarbonate type (HCO₃⁻ varies between 2 and 7 meq/L) and the majority of water samples of the area comes under Normal Chloride type and remaining Normal Bicarbonate type. On the basis of base exchange indices, r₁ (r₁ = Na⁺ − Cl⁻/SO₄²⁻) and r₂ (r₂ = K⁺  + Na⁺ − Cl/SO₄²⁻) (Soltan 1999), water samples come under Na⁺–SO₄²⁻ type (r₁ > 1) and shallow meteoric percolation type (r₂ > 1) and the details are compiled (Table 10 and Fig. 4a, b). As Na⁺/Cl⁻ molar ratio is > 1 in many samples, it may be noted that silicate weathering is responsible for Na⁺ into the sample (Meybeck 1987).

Table 10 Different parameters of premonsoon water samples, 2013

Fig. 4

a Base exchange index (r₁) plot—PRM, 2013. b Meteoric genesis index (r₂) plot—PRM 2013

4.1 Modified Piper Diagram

The groundwater samples of PRM, 2013 plotted in a Modified Piper diagram (Chadha 1999) to understand the water type and hydrochemical nature of the samples. It is revealed that the groundwater type in the area mainly belongs to Na⁺–Cl⁻ type (Fig. 5).

Fig. 5

Modified piper diagram—PRM, 2013

The deeper waters of the area generally with alkali metals surpass alkaline earths, strong acidic anions exceed weak acidic anions, (Na + K) > (Ca + Mg) > (Cl + SO₄) > (CO₃ + HCO₃) but the Veliyanad sample (Na–HCO₃ Type) with Alkali metals surpass alkaline earths and weak acidic anions exceed strong acidic anions, (Na + K) > (Ca + Mg) > (CO₃ + HCO₃) > (Cl + SO₄). The diagram indicates that Na⁺ and Cl⁻ and/or HCO₃—predominate in the samples. The dominance of Na⁺ may be due to a negative index of base exchange (Pophare and Dewalkar 2007). The Na⁺ and Ca²⁺ are in a transitional state with Na⁺ replacing Ca²⁺ and HCO₃⁻, Cl⁻ due to physico-chemical changes in the groundwater reservoir and the water–rock action. Because of cation exchange, Na⁺–HCO₃⁻, Ca²⁺–Cl⁻, and Na⁺–SO₄²⁻, the groundwater types may form according to cation exchange capacity of the matrix; and is a vice versa reaction (Elango and Kannan 2007). This cationic exchange is the principal source of sodium and the water in the deeper aquifer is seawater type and base exchange water is only of restricted type.

4.2 Hydrogeochemical Evaluation

The high sodium content among cations in the groundwater for the period could be due to weathering of silicate rocks which was further enhanced by evaporation and/or consumptive use. When Na⁺/Cl⁻ molar ratio will be 1, halite liquefaction is answerable for sodium dominance and > 1, Na⁺ is discharged from silicates by weathering activity (Meybeck 1987). The Na⁺/Cl⁻ molar ratio is > 1 in the samples of water can change to brine rich in NaCl if it meets highly water-soluble chloride minerals, in contact with evaporates (Gosselin et al. 2003).

The study of the Ca²⁺/Mg²⁺ ratio shows that the liquefaction of silicates was one of the processes involved in attaining the present chemical setup of the groundwater. The Ca²⁺/Mg²⁺ ratio of 1 point action of dolomite and > 2 effects of silicate minerals on the hydrochemistry; recommended Calcite activity for Ca²⁺–Mg²⁺ distribution in groundwater (May and Loucks 1995). As the Ca²⁺/Mg²⁺ ratio is between 1 and < 2, dolomite dissipation is accountable for Ca²⁺–Mg²⁺ contribution (Table 10). The scatter plot of Fig. 6 reveals that the bulk of samples fall below the equiline, pinpointing silicate weathering was the essential process involved in the genesis of groundwater (Datta and Tyagi, 1996). The plot of Ca²⁺ + Mg²⁺ versus HCO₃⁻ + SO₄²⁻ confirms the influence of dolomite dissolution on Ca²⁺–Mg²⁺ in the water.

Fig. 6

The scatter diagram of Ca²⁺ + Mg²⁺ versus HCO₃⁻ + SO₄²⁻

4.3 Gibbs Plots, 1970

The genesis of groundwater and the role of water–rock interaction can be understood by plotting TDS versus Na⁺/(Na⁺ + Ca²⁺) for cations and TDS versus Cl⁻/(Cl⁻ + HCO₃⁻) for anions . It is disclosed that the water samples, by and large, fall in the evaporation predominance to rock predominance. No the basis of the falling of the plots, it can be suggested that evaporation played a better role in water chemistry and water–rock interaction minor role (Fig. 7a, b). It is also to be noted that geological localization also plays a dominant role in groundwater hydrochemistry (Beck et al. 1985).

Fig. 7

a Gibb’s plots for cation—Pre-monsoon, 2013. b Gibbs plots for anion—PRM, 2013

4.4 Chloroalkali Indices

The role of aquifer material in the evolution of groundwater was analysed by calculating the Chloroalkali indices for cations (CAI-1) and anions (CAI-2). The CAI-1 [Cl⁻ − (Na⁺ + K⁺)] Cl⁻ and CAI-2 [Cl⁻ − (Na⁺ + K⁺)/(SO₄²⁻  + HCO₃⁻ + CO₃⁻ + NO⁻)], after Schoeller (1967), associate the ion exchange between groundwater and aquifer material. The CAI-1 and CAI-2 with negative values point to ion exchange between Na⁺−K⁺ in water and Ca²⁺–Mg²⁺ in rocks (McIntosh and Walter 2006).

4.5 Irrigation Suitability

The irrigation suitability has been tested by various tools, methodology and the analytical results are compiled (Table 11 and Fig. 8).

Table 11 Methodology for estimation of irrigation usefulness

Fig. 8

Wilcox classification of irrigation water

The EC is a measure of salinity hazard to crops and is classified into five major types, as per Raghunath and that 5 samples in the study area are in a good category, 20 samples (80%) are within the permissible limits and can be used for irrigation on almost all types of soil and 2 samples are in a doubtful class.

The Sodium Absorption Ratio, SAR is a sign of sodium hazard for irrigation purposes (Gholami and Srikantaswamy 2009). The 12, 48, and 36% of the samples are excellent, good, and doubtful types, respectively, and that at Arookutty is unsuitable (Richards 1954). The plotting of the data in Wilcox’s diagram reveals that the bulk of samples come under the doubtful to the unsuitable category of per cent Sodium and the EC permissible to doubtful (Fig. 8).

The elevated per cent sodium, % Na can enhance the exchange of Na⁺ in irrigated soil and change soil permeability, and structure and make a toxic situation for flora (Bangar et al. 2008; Durfer and Backer 1964). It is found that 8% of samples were doubtful and all others unsuitable for irrigation. As per the permeability index, PI (Doneen 1964) majority of the sample (92%) is apt for irrigation in all types of soil. Kelley (1940) and Paliwal (1967) projected the appropriateness of irrigation water quality based on the Na⁺ against Ca²⁺ and Mg²⁺. As Kelley’s Index (KI) values are more than 2, the water in the area is not fit for irrigation. Here soluble sodium percentage and SSP values are more than 50, hence water is not suitable for cultivation. As all the samples with Magnesium Ratio, MR values less than 50, the water is good for agricultural practices (Lloyd and Heathcote 1985) (Table 12).

Table 12 Hydrochemical characteristics (PRM 2013)

5 Conclusions

The predominance of ions in the groundwater of the deep coastal aquifer of Alappuzha, Kerala, India is in the order of Na⁺ ≥ Ca²⁺ ≥ Mg²⁺ ≥ K⁺ The Factor analysis has been applied to extract the components which influence the water qualities. The three factors were extracted. Factor 1, which is associated with the variables EC, TDS, Cl⁻, and Na⁺ explained 27.2–37.1% of the variance. Factor 2 accounts for 22.4% to 25% of the variance. Factor 2 has high loadings with T_H, Fe⁺⁺, Ca⁺⁺, Mg⁺⁺, HCO₃⁻, K⁺, and F. Factor 3 accounts for 10.2% to 13.2% of the variance, which is associated with pH, SO₄⁻, and NO₃. Factors 1 and 2 pinpoint natural processes and water–rock interactions. The NO₃ and SO₄⁻ of Factor 3 are affected by fertilizers and geologic formations. This coastal tract is also indicative of the high salinity with high sodium content in water. The higher pH and higher concentrations of TDS, Cl⁻, Fe⁺⁺, and fluoride may result in various health hazards and the use of the water of tube wells should be restricted or appropriate de-flouridization techniques should be incorporated in the drinking water supply schemes in the district. Since groundwater finds domestic use various types of rainwater harvesting especially tube well recharge should also be adopted in the area. The detailed study on chemical evolution revealed that the waters in the area are of seawater type, normal chloride type with evaporation played the main role in the chemistry of water followed by water–rock contact and groundwater reservoir. The irrigation suitability was tested by various tools and it was found that in many of the areas, the fluoride-enriched waters are not suitable for agricultural practices.